RECYCLED CONSTRUCTION AND DEMOLITION MATERIALS IN PAVEMENT AND BASES

Arul Arulrajah1, Suksun Horpibulsuk2 and Farshid Maghool3

1,3Department of Civil and Construction Engineering, Swinburne University of Technology, Australia; 2School of Civil Engineering and Center of Excellence in Innovation for Sustainable Infrastructure Development, Suranaree University of Technology, Thailand

ABSTRACT The increase in generation of waste from human and construction activities along with significant increase in global population has led to major issues over waste disposal. The reuse of waste material is an important topic from both sustainability and economic point of view. In this paper, application of Construction and Demolition materials (C&D) in works is reviewed and discussed. C&D material were evaluated by laboratory testing methods to assess their viability for reuse in and . Several unique field case studies where C&D materials have been used are also reported. The environmental risks associated with use of C&D materials in road is also discussed. The types of Construction and demolition wastes that were studied include Recycled Aggregate (RCA), Crushed (CB), Reclaimed Pavement (RAP), Fine Recycled Glass (FRG) and Waste Rock (WR). C&D materials were found to be suitable for road and footpaths applications such as embankment fills, pavement base/subbase and pipe bedding applications.

Keywords: Waste Materials, Pavement, Base, Subbase

INTRODUCTION impact [9]. Removing obstacles for the reuse of C&D Waste materials are any type of material materials in road work applications will have a by-product of human and industrial activity that has profound impact in moving toward a more no lasting value [1]. The disposal of solids waste is a sustainable global environment. major problem throughout the world. The sustainable In this research the geotechnical characteristics of usage of waste materials in engineering applications five major categories of C&D materials, and several is of social and economic benefit to all nations. Due of their blends, have been characterized through an to the shortages of natural mineral resources, extensive series of geotechnical tests to address their available land space and increasing waste disposal usage in road pavement and footpath applications. costs, recycling and reuse of solid wastes has become Field tests results from several unique case studies significant in recent years. utilizing C&D materials are also presented. The Construction and Demolition (C&D) materials are properties of the C&D materials were tested and the excess or waste materials associated with the compared with typical road authority specified construction and demolition of buildings and requirements for usage as a subbase material. structures, including concrete, brick, reclaimed Environmental assessments have also been carried asphalt, steel, timber, plastics and other building out. The C&D materials studied in this research were materials and products [2]. Recycling and subsequent Recycled Concrete Aggregate (RCA), Crushed Brick reuse of C&D materials will reduce the demand for (CB), Reclaimed Asphalt Pavement (RAP), Waste scarce virgin natural resources and simultaneously Rock (WR) and Fine Recycled Glass (FRG). reduce the quantity of this waste material destined for landfills [3]-[5]. The usage of C&D materials in civil RCA is the by-product of construction and engineering applications such as road and footpaths is demolition activities of concrete structures [2]-[10]. a viable and sustainable option to minimise the C&D CB is a by-product of demolition activities of waste while reducing the demand for scarce virgin buildings and other structures [11]. WR used in this quarried materials [1]-[5]-[8]. study originates from ―basalt floaters‖ or surface In Australia, approximately 8.7 million tons of excavation rock (basalt) which commonly occurs near RCA, 1.3 million tons of CB, 3.3 million tons of WR, the surface to the west and north of Melbourne, 1.0 million tons of FRG and 1.2 million tons of RAP Australia [12]. RAP is the name given to spent are stockpiled annually [8]. Reuse of C&D materials asphalt that has been recycled during removal from would clearly provide substantial benefits in terms of roadways which is done on a regular basis [13]. FRG reduced new material supply and waste disposal cost, is the by-product of crushing mixed color bottles and increased sustainability, and reduced environmental other glass products collected from both municipal and industrial waste streams [14].

1 PREVIOUS STUDIES A study on recycling and reuse of brick in United Kingdom was undertaken by Reference [20]. Their Recycled Concrete Aggregate (RCA) study discussed UK’s current brick recycling strength In an experimental research work performed by and proposed new brick recycling technology to Reference [15] on crushed concrete aggregate, the achieve higher economic and environmental researchers emphasized that the findings of research performance. on one demolition waste should not be applied to other recycled materials, as many different types are Reclaimed Asphalt Pavement (RAP) produced [15]. Reference [21] conducted a laboratory evaluation Reference [1] investigated the physical of cement stabilized RAP and RAP-virgin aggregate characteristics of variable grades of recycled blends as an alternative for base layers. Test results aggregates and reported that the larger the size of the suggested that optimum moisture content, maximum aggregate, the smaller the percentage of cement dry density and strength of RAP increases with the mortar attached to its surfaces and the better the addition of virgin aggregate and cement as a stabilizer aggregate quality will be. Reference [1] concluded [21]. Test results suggested that pure RAP aggregate that recycled concrete aggregates have a larger can be utilized as a conventional base material only if amount of porosity and can potentially undergo a stabilized with cement [21]. The ability of RAP higher degree of deformation. aggregate to function as a structural component in Reference [6] conducted a study on use of crushed road pavements is more pronounced when it is concrete in road subbase layers and concluded that stabilized with cement rather than when blended with recycled coarse and fine aggregates of different virgin aggregate [21]. nominal sizes conform to the required grading limits Reference [5] carried out tests on RAP materials specifications for pavements, embankments, roads treated with different percentages of Portland cement and [6]. and with alkali-resistant glass fibers [5]. Test results Reference [9] in an experimental study found that confirmed the potential of cement-fiber-treated RAP recycled construction wastes have significant shear material as an environmentally and structurally sound strength which makes these materials an alternative to alternative to non-bonded materials in base and natural aggregate in various geotechnical applications. subbase layers of road pavements [5]. The authors reported reductions in the frictional Reference [7] conducted a series of repeated load resistance of these materials caused by repeated triaxial tests in a research study to evaluate the loading [9]. Reference [16] studied the shear strength effectiveness of adding cement in enhancing resilient behavior of recycled construction materials for characteristics of RAP aggregates with promising projected use in vibro-ground improvement findings on cement stabilized RAP. applications. It is found that for both dry and wet material, the drained internal angle of friction is Fine Recycled Glass (FRG) approximately 39°, which reduces to 32° when the There are several research publications available recycled concrete was mixed with clay slurry [16]. on using recycled crushed glass in concrete mixtures Reference [17] and Reference [18] studied the [22]-[23] and also in asphalt aggregate as a permanent deformation characteristics of RCA, RAP replacement to sand and gravel material [14]-[22]. and a dense-graded aggregate by conducting cyclic Recycled glass has been also suggested in load triaxial tests and reported that RCA accumulated applications such as backfill material [24], the least amount of permanent strain out of the three embankment fills and in pavements [14]. materials [17]-[18]. Reference [19] studied the resilient modulus and permanent deformation of RCA Waste Rock (WR) and reported that the material was suitable for Reference [25] studied the behavior of sandstone unbound base courses. and shale aggregates under cyclic loading for the purpose of using them in unbound forest roads. Test Crushed Brick (CB) results suggested that sandstone had very good Reference [15] conducted an experimental study resistance to deformation and rutting while shale had on particle size distribution, water absorption, poor resistance [25]. flakiness index, particle density, compaction Reference [26] studied the mechanical behavior of characteristics, aggregate impact and aggregate treated crushed rock used in road base layers through crushing value of CB concluding that crushed brick a range of static and repeated load triaxial tests. had significantly different engineering properties to Reference [27] conducted a research on using marble crushed concrete. and andesite quarry wastes in asphalt pavements. Reference [6] investigated the possibility of using Test results implied that physical properties of these CB aggregates in unbound subbase layers in Hong waste aggregates are within specified limits and Kong and noticed the inferior shear performance of consequently they can potentially be used as crus hed brick in CBR tests compared to RCA. aggregates in light to medium trafficked asphalt pavement binder layers [27]. 2

2 Reference [28] examined the shear behavior of 40 surcharge mass of 4.5 kg was placed on the surface of mm uniform crushed recycled rock (quarry waste) in the compacted specimens and then the samples were a study of their use in ground improvement works in soaked in water for a period of four days. This is to the UK. Test results suggested that the presence of simulate the confining effect of overlying pavement slurry has adverse effects on shear strength and layers and also the likely worst case in-service settlement potential of quarry waste aggregates [28]. scenario for a pavement Reference [34]. The static Reference [29] carried out a research use of triaxial tests were performed in an automated triaxial secondary materials for pavement construction in the testing system with specimen dimensions of 100 by UK and undertook a range of tests including repeated 200 mm (diameter by height) for all recycled material load triaxial test on mine-rockwaste and slate waste. types except FRG which was tested with the dimensions of 50 by 100 mm. The test specimens LABORATORY EVALUATION were compacted to 98% of MDD from modified compaction test in a split mold in eight layers. The Materials and methods compaction was done by mechanical compactor with C&D aggregates were obtained from several around 15 blows of modified compactive effort of recycling sites in the state of Victoria, Australia. The 2700 kN-m/m3 for each of the eight layers. Triaxial recycled CB, RCA, WR and RAP used in this compression (shearing) was executed on the saturated research had a maximum particle size of 20 mm. FRG and consolidated specimens. The samples were has a maximum particle size of 4.75 mm and compressed at the given consolidated confining comprises of sand size and a small percentage of silt pressures under drained conditions (CD test). The size particles Reference [4]. During sampling; ASTM shearing was performed under strain-controlled practice for sampling aggregates was carefully condition at the selected strain rate of 0.01 mm/min. practiced and all necessary precautions were taken to Replicate samples were tested for the triaxial tests at capture a sample containing representative particle the various stress levels. sizes and all contaminants. Repeated load triaxial (RLT) tests were conducted Laboratory tests were subsequently undertaken on to determine the resilient modulus and permanent these recycled C&D aggregates, and several of their deformation of the recycled materials. In this blends. The laboratory investigation included basic investigation, the RLT test was performed according characterization tests such as particle size distribution, to Austroads Repeated Load Triaxial Test Method particle density (coarse and fine fraction) and water AG: PT/T053 [35]. The RLT testing consists of two absorption (coarse and fine fraction), organic content, phases of testing, permanent strain testing and then pH, hydraulic conductivity, flakiness index, Los resilient modulus testing. Permanent strain testing Angeles (LA) abrasion, modified Proctor compaction consists of three or four stages, each undertaken at and CBR tests. Shear strength tests were subsequently different deviator stresses and a constant confining undertaken with static triaxial tests. RLT tests were stress. The resilient modulus testing consists of sixty undertaken to determine the permanent deformation six (66) loading stages with 200 repetitions. In this and resilient modulus characteristics of the C&D test, the specimens were compacted to 98% MDD materials. The room temperature was maintained at based on modified compaction effort and tested at 20±1°C for the triaxial and RLT tests. three target moisture contents of 70%, 80% and 90% Using sieve analysis results, Unified Soil of the OMC based on modified compaction effort, so Classification System [30] was implemented to as to simulate the dry-back process in the field. classify the recycled materials. Organic content of all Replicate samples were tested for the RLT tests at the recycled material sources in this research was each of the various moisture levels. determined following Reference [31]. pH values of Total Concentration (TC) and leachate analysis the recycled materials were determined following the were carried out for a range of heavy metals on Standards Australia [32]. samples of C&D material. In the preparation of The test specimens for hydraulic conductivity leachate, the method described in Australian Standard tests were compacted with modified Proctor was followed and slightly acidic leaching fluid (pH = 5) and alkaline leaching fluid (pH = 9.2) were used as compaction effort, at optimum moisture content (OMC) to reach at least 98% of maximum dry density leaching buffers [36]. (MDD). The falling head test method was chosen for all recycled aggregate with the exception of FRG Results and Discussions which was tested by the constant head method. The geotechnical properties of the various The flakiness index tests were carried out recycled C&D materials and discussions on the following Reference [33]. Oven dried samples that laboratory evaluation results are presented in this passed 63.0 mm and retained on the 6.30 mm were section. selected for testing.

CBR test specimens were prepared by applying modified compaction efforts to recycled aggregates mixed at the OMC obtained in compaction tests. A 3

3

Classification, index and geotechnical properties The particle size distributions of the five recycled materials are shown in Figure 1. The ―after compaction‖ grading curves show that some breakdown has occurred during compaction especially for CB and RAP material. However, all the recycled C&D materials, except for FRG, satisfied the guidelines for type 1 gradation C road base material according to ASTM D1241-07, except for slight deviations in the finer side for some materials. The grain size distribution parameters including D10, D30, D50, D60, Cu, Cc, percentage of gravel, sand and fine particles, USCS symbol and description are summarized in Table 1. RCA, CB, WR, and RAP have approximately Figure 2. Modified compaction curves of C&D equal amount of sand and gravel sized fractions, materials [8]. enabling them to be classified as well-graded gravelly sand or sandy gravel. The pH values of all blends are above 7 and this The fine fractions used for Atterberg limit tests indicates that the blends are alkaline by nature. (i.e. particles smaller than 0.425 mm) are low or non- Hydraulic conductivity of the recycled materials plastic and are mainly sand or silt by nature, so the ranges from 1.75×10-5 to 4.50×10-9 m/s. These plastic limit and liquid limit could not be obtained for values can be described as low permeability for RCA any of C&D material studied in this research. and CB and high permeability for WR, RAP and FRG. The hydraulic conductivity values of RCA and CB are lower and WR, RAP and FRG are higher than that of 6.59×10-8 m/s of reported for natural aggregate with similar classification [6]. Flakiness index is relevant for aggregates used in bituminous mixtures. The flakiness index values for the recycled materials varied from 11 to 23. This is however still within the requirements of typical state road authorities for usage as a base material, which specifies a maximum value of 35. Reference [1] also suggested 40 as the flakiness index upper limit for aggregates to be used in pavement applications. Flakiness index values are not relevant for the FRG Figure 1. Particle size distributions of the five as flakiness index is not applicable to material recycled materials before and after compaction [8]. passing 6.30 mm sieve. Particles crushing and degradation is considered Figure 2 shows the modified compaction curves as a significant issue in certain geotechnical of C&D materials which possess characteristic applications and accordingly any attempt to utilize convex shaped curves similar to natural aggregates recycled materials in geotechnical engineering [24]. The modified compaction test results indicated applications should examine this issue carefully [9]. that WR had the highest MDD while FRG had the An LA abrasion maximum value of 40 is normally lowest value. The fact that FRG indicated the lowest adopted by state road authority specifications for MDD is consistent with the finer gradation curve of pavement subbase materials [34]. RCA, WR and FRG and its lower particle density for both fine and FRG meet this maximum criteria, CB is just below coarse fractions. The flatter compaction curve of FRG the limits while RAP with a value of 42 is above the suggests its low sensitivity to water content changes limits. This indicates that RCA, WR and FRG are in comparison to natural aggregate which gives FRG more durable in abrasion than CB and RAP. This stable compaction behavior and good workability [4]- further substantiates that the gradation curves of CB [24]. The OMC of the C&D materials indicated that and RAP showed the highest change after modified RAP had the lowest OMC of 8.0% while CB had the compaction tests. This further suggests that RAP may highest of 11.25%. have to be blended with other aggregates to enable its Organic contents were found to be low for the usage in pavement subbase applications. The abrasion recycled materials and high for the RAP. This could loss value obtained for RCA in this study is very be due to the presence of bitumen in RAP that is rich close to the value of 25 for a recycled concrete in carbon. investigated by Reference [37].

4

4

Table 1. Geotechnical Properties of Recycled C&D Materials [38].

Typical quarry Geotechnical parameters RCA CB WR RAP FRG material

D10 (mm) 0.24 0.18 0.075 0.24 0.16 — D30 (mm) 1.3 1.7 1.5 1.9 0.45 — D50 (mm) 5.0 5.6 3.9 4.5 0.85 — D60 (mm) 7.5 8.0 5.6 5.9 1.2 — Cu 31.2 44.4 74.7 25.6 7.5 — Cc 0.9 2.0 5.4 2.5 1.5 — Gravel content (%) 50.7 53.6 44.7 48.0 0.0 — Sand content (%) 45.7 39.8 45.1 46.0 94.6 — Fines content (%) 3.6 6.6 10.2 6.0 5.4 <10 USCS classification GW GW SW GW SW — Particle density—coarse 27.1 26.2 28.1 23.5 24.4 >19.62 fraction (kN/m3 ) Particle density—Fine fraction 26.0 25.8 28.0 23.4 24.3 >19.62 (kN/m3 ) Water absorption—coarse 4.7 6.2 3.3 2.2 1.0 <10 fraction (%) Water absorption—fine fraction 9.8 6.9 4.7 2.4 1.8 <10 (%) MDD (kN/m3 )—modified 19.13 19.73 21.71 19.98 17.40 >17.5 compaction OMC (%)—modified 11.0 11.25 9.25 8.0 10.5 8–15 compaction Organic content (%) 2.3 2.5 1.0 5.1 1.3 <5 pH 11.5 9.1 10.9 7.6 9.9 7–12 Hydraulic conductivity (m/s) 3.3 × 10-8 4.5 × 10-9 2.7 × 10-7 3.5 × 10-7 1.7 × 10-5 >1 × 10-9 Flakiness index 11 14 19 23 — <35 LA abrasion loss (%) 28 36 21 42 25 <40 CBR (%) 118–160 123–138 121–204 30–35 42–46 >80 Triaxial test (C&D): apparent 44 41 46 53 0 >35 cohesion (kPa) Triaxial test (C&D): friction 49 48 51 37 37 >35 angle (degree) Resilient modulus: target 90% 239–357 301–319 121–218 — — 125–300 of the OMC Resilient modulus: target 80% 487–729 303–361 202–274 — — 150–300 of the OMC Resilient modulus: target 70% 575–769 280–519 127–233 — — 175–400 of the OMC

Shear strength properties Consolidate Drained triaxial tests undertaken on A CBR value of at least 80% is typically the recycled materials indicated that RCA, CB, WR required by state road authorities for a subbase had a drained cohesion ranging from 41 kPa to 46 material in Victoria [8]. RCA, CB and WR meet the kPa and a drained friction angle ranging from 49° to CBR requirements for usage as a subbase material. 51°. This indicates the shear strength parameters for However, FRG and RAP would need to be blended these recycled materials are in the range of coarse with other aggregates to improve their CBR aggregates. FRG had a drained cohesion of 0 kPa performances to become suitable for road subbase which indicate the properties of FRG are similar to layers. These two recycled materials are however coarse sand with little to no cohesion. RAP and FRG suitable for usage as a fill material in embankments, had similar low drained friction angles of 37°, which need far lower CBR requirements. similar to that of loose sand.

5 WR in this study originates from basalt floaters or basalt surface excavation rock which is commonly found during subdivision and excavation works. When these waste materials are excavated and disposed, they are disposed together with excavated fine materials which contribute to high cohesion values for the WR material. Furthermore, the addition of water to the WR during the compaction to the OMC could result in the fines present forming a paste which subsequently contributes toward a higher cohesion value. RCA comprises of a high amount of cement dust and fines. Cementing and bonding could result when Figure 3. Results of resilient modulus for RCA, CB water is added to the crushed concrete when the and WR [8]. samples are compacted to the OMC and MDD. Unreacted cement in the crushed concrete would Fill material consists of soil (being clay, silt react with water to provide cohesion and this would and/or sand), gravel and rock of naturally occurring result in the high cohesion noted from the crushed materials and is often referred to as clean fill by concrete in the triaxial shear tests. industry, and may be suitable for site filling or The RLT test provides resilient modulus– leveling depending on an assessment of contaminant permanent deformation parameters that uniquely levels and intended use [40]. Soil and aggregates describe the material response to loading may be classified as fill, when an assessment under prevailing physical conditions. These demonstrates that the material is not contaminated or parameters are used as input to the design and the contamination levels in form of TC are not analysis of pavement structures [39]. The test results higher than the values specified as maximum TC for are used to establish a material selection criterion fill material [40]. based on its ability to perform effectively in terms of TC values of C&D samples were compared with permanent deformation sustained. Results of EPA Victoria (Australia) requirements for fill resilient modulus testing for RCA, CB and WR are material. The comparison implies that for all the presented in Figure 3. contaminant constituents with the exception of The RLT test results indicated that RCA, CB and chromium, TC values of C&D samples are far below WR performed satisfactorily at 98% MDD and at a the threshold. The chromium metal is found in a few target moisture content of 70% of the OMC. RCA, oxidation states such as hexavalent chromium CB and WR materials showed sensitivity to moisture (chromium VI) and trivalent chromium (chromium and produced higher limits of permanent strain and III). The values reported for C&D samples are the lower limits of resilient modulus, particularly at total chromium (chromium III + chromium VI) higher target moisture contents in the range of 80%- while the EPA Victoria requirement is on hexavalent 90% of the OMC. The performance of RCA, CB and chromium (chromium VI). As such, C&D materials WR were found to be affected by increasing the will go beyond the chromium boundary only and target moisture contents and the density level. This only if all the chromium found in the test is of type is apparent particularly for CB which failed at the chromium VI which does not seem to be the case higher target moisture contents of 80%-90%. The here [4]. results of permanent strain and resilient modulus for According to the US EPA, a material is RAP and FRG could not be reported as these two designated as a hazardous waste if any detected materials possess very low cohesion values and their metal occurs at concentrations larger than 100 times samples failed within a few cycles at a target the drinking water standard [24]. The ASLP values moisture content of 60% of the OMC. Consequently, were found to be below the threshold of hazardous the tests for higher target moisture contents were not waste proving that they will not be categorized as attempted for RAP and FRG. RCA and WR have hazardous waste according to U.S. EPA. much smaller permanent strain and much higher modulus than natural granular subbases, which FIELD CASE STUDY 1: RECYCLED GLASS indicate their performance as superior or equivalent IN PAVEMENT BASE to typical quarry subbase materials. Materials and Methods Total concentration and leachate tests Samples of recycled aggregates were obtained Using the method described previously, ASLP from a recycling site in the state of Victoria, tests with two buffer solutions (acidic and alkaline) Australia. FRG has the maximum particle size of were conducted on representative samples of C&D 4.75 mm while the other recycled aggregates materials. (WR, RCA) had a maximum particle size of 20 mm.

6 FRG has also a lower compacted density compared was conducted at various locations after the to RCA and WR. The RCA and WR are stiffer, more placement of the pavement base layers. The field durable and readily accepted materials in pavement tests were undertaken 3 days after the placement of construction. FRG was researched as a the subbase and base layers with a Nuclear Density supplementary material in limited combinations with Gauge (NDG) and Clegg Hammer (CH). It was the more robust and readily accepted RCA and WR expected that the field moisture conditions at the materials. Laboratory tests were undertaken on these time of testing would be lower than the optimum various recycled aggregates by using specified moisture conditions at the time of compaction, as the international standards. materials were delivered within the recycling site Nine sections of unbound granular base and haulage time was 1-2 minutes. pavements, comprising of up to 30% Fine Recycled Glass (FRG) in blends with Recycled Concrete Aggregate (RCA) and Waste Rock (WR) in the pavement base were constructed on the main haul road at a recycling site in Melbourne, Australia. Each of the pavement sections was 80 m in length and 4.75 m in width. The design of these granular base pavements was based on the outcomes of the initial laboratory testing phase of this research. FRG/RCA and FRG/WR blends were found to satisfy the requirements of a pavement subbase material in the laboratory testing phase of this research, however it was decided to use this material in the pavement base and assess its performance as a higher quality pavement base material in the field Figure 4. Laying of the base layer at one of the trials. The 200 mm thick base layer comprised of sections including a fraction of FRG [38].

FRG blends was placed and subsequently overlaid Results and Discussions by a 50 mm thick glass asphalt (glassphalt) cover Direct transmission method of nuclear density comprising of asphalt with 5% glass content. The and moisture testing was conducted on the granular base materials composition was with blends bases after the construction of each layer at 10 meter comprising of 10% to 30% content of FRG/WR or intervals along 2 wheel paths for each of the FRG/RCA. pavement sections. Field density values were Two control sections comprising 100% of WR calibrated by using oven moisture tests obtained and RCA were also built. These aggregates are from the same locations as moisture contents commonly accepted for usage in pavement base attained by using the nuclear gauge. Samples of base applications in Australia. Four sections were materials being FRG/RCA and FRG/WR blends constructed with RCA with 10%, 15%, 20% and were obtained from the base layers from each 30% of FRG. Another three sections were section placed on construction and subsequently constructed with WR with 10%, 20% and 30% of tested in the laboratory to obtain their corresponding FRG. Figure 4 shows the laying of the base layer at Maximum Dry Density (MDD) and Optimum one of the sections. Moisture Contents (OMC). The 200 mm thick pavement base layer was The average field densities of the FRG-WR constructed with various FRG/RCA or FRG/WR sections are noted to be higher than that of the blends in 7 sections and with RCA and WR for the 2 FRG/RCA blends. The FRG/WR blends also had remaining control sections. Each granular base higher field density and laboratory density results material was mixed to the appropriate optimum than corresponding FRG-RCA blends with the same moisture content in the pugmill at the recycling site glass contents. The results indicate that average and immediately transported by truck to the site. A density ratios in individual sites varied in the range minimum 3 days dry-back period was applied to of 96% to 100% MDD. The Control Sections (RCA, each lift. During the dry back periods, Nuclear WR) as expected achieved the highest density results Density testing was conducted for each lift to check in the field and laboratory density tests as compared density and moisture content. Nuclear Density tests to the FRG/RCA and FRG/WR sections. The field were also conducted to measure the final compaction results indicate that WR is a higher quality material levels of the combined 200 mm base. Final levels of than RCA. the base surface were also taken to confirm base It was noted that the FRG30/WR70 blend thicknesses. containing 30% recycled glass content produced the For the assessment of the geotechnical lowest density ratio compared to other FRG/WR performance of the recycled materials and their blends with less glass additive content. Similarly, the impact on base strength and stiffness, field testing FRG30/RCA70 blend containing 30% recycled glass

7 content also produced the lower density ratio compared to other FRG/RCA blends. From this finding, it can be concluded that, a blend containing a recycled glass additive content of greater than 20%, would likely result in a lower field dry density being achieved. Road authorities require material to have minimum mean values of density ratio of 100% for base materials for light duty pavements. The base layers were also found to be marginally below these requirements except for the RCA and FRG15/RCA85 sections. Road authorities also require material during compaction to have a moisture content of not less than 85% of optimum Figure 5. Clegg Hammer results for various during compaction and, after completion of combinations of C&D pavement base sections [38]. compaction of a layer. The field moisture content results indicated that FIELD CASE STUDY 2: RECLAIMED for the FRG/RCA blends, the average moisture ASPHALT PAVEMENT IN PAVEMENT contents varied in the large range of 6-8.3% which is SUBBASE due to variation in their respective OMCs and differing dry back times. The overall moisture Materials and methods variations were higher than the average moisture RAP was used as a subbase material in nine contents of the FRG/WR blends, which varied pavement sections for a haul road at a recycling site within a much smaller range of 4.9 to 5.2%. The operator’s facility. RAP was selected as it was control sections (RCA, WR) had field moisture available in large stockpiles at the recycling site and contents of 6.9% and 5% respectively, which were there was interest from various parties to evaluate fairly consistent with the FRG/RCA and FRG/WR the field performance of untreated RAP in subbase blends. layer. The results from the Clegg Hammer tests were Each of the pavement sections was 80 m long analyzed to determine CBR values of the various and 4.75 m wide. The pavement sections comprised pavement sections as well as to determine the of a 200 mm thickness RAP subbase, overlying a strength ratios after field compaction. Figure 5 subgrade with a design soaked CBR greater than 5%. presents the Clegg Hammer results for CBR for the After placement and spreading, the RAP material various pavement base sections that have been was graded to a uniform level using the controlled transposed into the same figure for easy reference. grader. A minimum 4 days dry-back period was Results of Clegg Hammer tests meet the minimum applied for the RAP subbase in all pavement soaked field CBR of 100% for base materials in all sections. Nuclear density checks were undertaken sections except in FRG10/WR90, FRG10/RCA90, during the dry back period to measure the final FRG20/RCA80 and FRG30/RCA70. Also, Clegg compaction levels of the RAP subbase. Final levels Hammer tests Results meets the specified minimum of the subbase surface were also taken to confirm soaked field CBR of 80% for subbase materials in subbase thicknesses. all sections except over short stretches of For the assessment of the geotechnical field FRG10/WR90, FRG20/RCA80 and FRG30/RCA70, performance of the RAP and their impact on subbase for 10 to 20 m in which it was marginally below the strength and stiffness, field testing was conducted at specified requirements but is still deemed acceptable various locations after the placement of the RAP for haul roads. pavement subbase layer using a Nuclear Density The results seem to indicate that FRG blends Gauge and Clegg Hammer 3 days after the should be limited to pavement subbase applications placement of the subbase layers. and may not meet requirements for a pavement base material. The Clegg Hammer results seem to also Results and Discussions indicate variation in the recycled blends within each The earlier phase of laboratory evaluation of pavement section and between pavement sections. RAP indicated that it did not meet the local road Both RCA and WR satisfied the requirements as a authorities’ requirements for usage in pavement subbase material. Limited blends of 20% FRG with subbase layers, particularly in terms of RLT and coarse sized recycled concrete aggregates CBR requirements. Nevertheless, the field trial (FRG20/RCA80) and waste rock aggregates, pavement constructed was for a private haul road in (FRG20/WR80), appears to be the optimum limits of the recycling operator’s site and as such did not have glass additives with recycled aggregates based on to meet the specified requirements of the local road the field testing results. authorities.

8 8 Direct transmission method of nuclear density The shared path comprised of a base layer of and moisture testing was conducted on the granular nominal 100 mm thickness, overlying a subgrade base and subbase layers after the construction of with a design soaked CBR greater than 3%. each layer at 10 meter intervals. Field density values The base layer was subsequently overlaid by a 30 were calibrated by using oven moisture mm thick asphalt cover. The asphalt footpath was measurements obtained from the same locations as constructed in three sections, with three different moisture contents attained by using the nuclear material blends of FRG/WR in the footpath base gauge. Samples of RAP were obtained from the layer. The three trial sections constructed were 15% subbase layers from each section placed on FRG section (FRG15/WR85) of 30 m length; 30% construction and subsequently tested in the FRG section (FRG30/WR70) of 85 m length and a laboratory to obtain their corresponding MDD and control section comprising basaltic WR of 125 m OMC. length. Figure 7 shows the FRG15/WR85 and The field densities for RAP were in the 20 FRG30/WR70 sections after completion of field kN/m3 range. The results indicate that density ratios compaction of base layer. in individual sites varied in the range of 94 to 97 % MDD. The results indicated that for the various sections, the average moisture contents obtained were in a similar range for the various pavement subbase sections varying from 4.9% to 5.5%. Road authorities require material to have minimum mean values of density ratio of 98% for subbase materials for light duty pavements. Based on this requirement, the RAP subbase layer was found to be marginally below these requirements. Road authorities also require material during compaction to have a moisture content of not less than 85% of optimum during compaction and, after completion of compaction of a layer. Based on the results in construction of the subbase for the RAP Figure 6. Clegg hammer test results for the pavement trial complied with target minimum various pavement subbase sections with RAP moisture content requirement of 85% OMC. [13] . The results from the Clegg Hammer tests were The local government council specifications for analyzed to determine CBR values of the various the state of Victoria, for aggregates in footpath bases pavement sections as well as to determine the were used to assess the geotechnical performance of strength ratios after field compaction. Figure 6 the recycled materials. The local government council presents the Clegg hammer results for CBR for the specifications for a shared footpath specifies: a various pavement subbase sections with RAP. The minimum laboratory soaked CBR of 40% (at 95% CBR values calculated from Clegg Hammer appear modified compaction); a maximum LA abrasion loss to vary significantly within each pavement section value of 60%; a maximum flakiness index value of and between pavement sections. 35% and a minimum soaked field CBR using the The Clegg Hammer tests indicate RAP did not Clegg Impact hammer of 28% at 92% modified meet the minimum soaked field CBR of 80% for compaction. In addition, there are recommended subbase materials in the various sections as advised particle size distribution upper (fine) and lower by local road authorities. Based on the field and coarse bound limits for the footpath base materials laboratory testing, RAP, had insufficient strength as shown earlier in Fig. 1. requirements to meet the local road authority The base material blends were mixed to the pavement subbase requirements. appropriate optimum moisture content in a pug-mill and transported by truck to the site. The material FIELD CASE STUDY 3: RECYCLED GLASS was planned to be placed and compacted at field IN FOOTPATHS moisture content close to the optimum moisture content, with the use of plant mixed wet mixes, to Materials and methods achieve a uniform density within the base. The final Laboratory tests were undertaken on prepared surfaces for all base materials were representative samples of FRG and WR and their regarded as very uniform. Field samples were blends. collected in sample bags from three locations in each An asphalt footpath for shared use by pedestrians section of the footpath base and combined into a and cyclists was constructed using the outcomes of single sample (>7 kg) for laboratory testing for the the laboratory testing phase of this research by a determination of compaction and particle size local government council in Melbourne, Australia. distribution properties.

9 93%, which are both greater than the minimum specified required relative compaction of 92% modified compaction. The variations between maximum and minimum dry density ratios along the chainage were found to be very small (i.e. maximum of 4%). This indicates the field densities were achieved very consistently along the chainage. The WR base in the trial shared path had a minimum relative compaction of 91%, which was on the borderline of the 92% minimum required relative compaction for base layer for modified compaction. It was noted that the density ratios between standard and modified compaction for these blends are in the Figure 7. A sections of shared path after field normal range of 94% to 96% of modified compaction typically obtained for quarried crushed compaction of base layer [38]. rocks. A comparison of these field samples with For the assessment of the construction variability samples tested during the initial laboratory of WR blended with FRG and their impact on base characterization phase indicates a very good match strength and stiffness, field testing was conducted at and confirms the quality of the recycled materials various locations along the pavement using a provided for the construction were consistent with Nuclear Density Gauge and Clegg Hammer after the those tested earlier in the laboratory characterization placement of the base layers. The tests were phase. Moisture content variations were high (i.e. conducted two days after the placement of the base maximum of 22%). This variation in the moisture layers. contents could be due to the exposure to sunlight and drying of the materials after compaction. Clegg hammer tests results were analyzed to Results and discussions determine CBR values of the various footpath Field samples of the footpath base materials were sections as well as to determine the strength ratios collected from each section during the footpath base based on a required minimum soaked field CBR of construction and sent for testing in the laboratory for 28% after field compaction. Figure 8 presents the particle size distribution and compaction tests. The Clegg hammer strength ratio results for CBR results from the testing of the field samples in this assessment of the various footpath sections. The phase were also compared to the earlier laboratory following assessments can be made based on the characterization phase results. field results obtained. In the laboratory characterization tests, gradation Clegg hammer tests on the FRG15/WR85 section curves were plotted for the materials before and after indicated field CBR values in the range of 30-52%. the compaction tests, to determine the degree of CBR values meet the specified minimum soaked breakdown of the particles, which was found to be field CBR of 28% using the Clegg hammer for all minimal. WR and FRG15/WR85 materials had a chainages. Strength ratios were in the range of 106- similar grading, whereas the FRG30/WR70 blend 184% (at the moisture condition of 66% of the with higher glass content have a grading exceeding optimum water content. the finer grading limit due to the higher glass content Clegg hammer tests on the FRG30/WR70 section of 30%. FRG possesses a gradation curve that indicated field CBR values were in the range of 25- exceeds the upper bound curves (finer grading 35%. CBR values meet the specified minimum limits), and this is the reason that FRG material soaked field CBR of 28% using the Clegg hammer could not be considered as a base material in the for all chainages, except Chainages 39-50, in which field trials. Furthermore, FRG may not meet it is just marginally below the specified requirements. requirements for workability during field Strength ratios were in the range of 88-126% (at the construction due to the same reason. All the various moisture condition of 76% of the optimum water material blends apart from FRG were found to plot content). within acceptable limits of the specified upper and Clegg Hammer tests on the WR section indicated lower limit requirements, except for some of the field CBR values in the range of 33-55%. CBR smaller fines at the lower particle sizes which were values meet the specified minimum soaked field considered acceptable. CBR of 28% using the Clegg hammer for all Comparisons were undertaken between the field chainages Strength ratios were in the range of 116- dry densities ratios obtained from the nuclear density 197% (at the moisture condition of 79% of the gauge with compaction results obtained from the optimum water content). field samples. FRG15/WR85 base had a minimum relative compaction of 94% and the FRG30/WR70 base also had a minimum relative compaction of

10 through Total Concentration and Leachate tests for a range of contaminant constituents particularly heavy metals and polycyclic aromatic hydrocarbons). These values then need to be checked with Environmental Protection Authority requirements of the country (or state) that material is going to be used at.

Minimizing possible environmental impacts Figure 9 shows the water flow balance in a recycled glass layer used in a subbase layer of a typical road pavement. It shows that part of the Figure 8. Clegg hammer results for CBR rainwater evaporates or flows off on top of the assessment of the various footpath sections [38]. surface and does not get into the pavement layer constructed out of C&D material [41]. The remaining part which is shown by gray arrows The FRG15/WR85 blend produced a high eventually seeps into the base layer and then fraction compaction ratio and moderately lower strength of it (indicated in solid black arrows) enters into the ratio. Therefore, adding 15% FRG to WR appears to recycled glass layer. From the recycled glass layer, significantly improve field workability, but the leachate will either move toward the drainage marginally reduces base strength. Between the pipe (and consequently will flow into surface water FRG30/WR70 blend and the other two sections, the streams or alternatively will be redirected into a FRG30/WR70 blend was found to produce a buffering zone) or it will seep directly into ground reasonable compaction value, but substantially lower water table. strength ratio. Therefore, finer grading (exceeding the fine grading limit) was found to significantly reduce the footpath base strength for FRG30/WR70. Limited blends of fine sized recycled glass with coarse sized rock aggregates, particularly FRG15/WR85, appears to be the optimum recycled material blend for a footpath base material.

ENVIRONMENTAL SUITABILITY

While there are several studies available on geotechnical properties of C&D material, few of them focus on road work applications and among these only a small number of them have discussed Figure 9. Water flow balance chart for a layer of the environmental concerns of using these materials recycled glass in road pavement [42]. in road applications [24]. This is despite the fact that one of the primary arguments against using recycled material including While in Figure 9 the cumulative widths of the C&D materials in road applications is the possible arrows indicate the rough proportions of the seepage spread of remaining pollutants [41]. Application of flow [42]; an appropriate design tries to minimize C&D materials in road works requires a the width of the arrow showing the percentage of comprehensive study on the environmental effects of leachate moving towards the groundwater. these materials to ensure that their environmental Appropriate design in road construction using impacts are considered throughout the life cycle of C&D materials including road pavements and the project [41]. embankment fills if required set targets in a way to minimize the percolation of mineral recycling materials. As a consequence, contaminant release Approach and methodology can be minimized or partly prevented. On the other Prior to using C&D materials in road work hand, the possible environmental impacts of using applications (such as embankment fills and C&D materials in road pavement applications can be pavement layers) all the possible environmental minimized by using these materials in places where risks including the leaching hazard, exposure of it is capped (for example beneath a contaminant constituents into soil, surface and surface as shown in Figure 9) or in locations that are ground water as well as the potential to spread into elevated above the ground level (such as in an surrounding areas during the service life of the ) [43]. project should be investigated. This is mainly done

11 Another appropriate design is to perform the CONCLUSIONS construction geometries and selection of soil materials in a way that the seepage water bypasses This paper has reported on a comprehensive C&D materials or its infiltration into layers of C&D laboratory evaluation of the properties of C&D materials is minimized. This will reduce the average materials. In addition, several unique case studies concentrations (averaged over the cross-sectional are also reported in this paper, comprising of actual area of the construction) at the bottom of such field implementation of recycled C&D materials in constructions. pavement bases, pavement subbases and footpath However, part of the water flow will eventually base applications. pass through the C&D material layer as shown in C&D materials were found to be suitable for Figure 9. If required and deemed necessary road and footpaths applications such as embankment designing a water purification plant connected to the fills, pavement base/subbase and pipe bedding recycled material layer with a drainage system, can applications. The sustainable usage of C&D help in diverting the leachate into the water materials in sustainable civil engineering purification plant. If the infiltration capacity of the applications will result in a lower carbon footprint subsoil is very low, the leachate might flow laterally for our future roads, footpaths and other civil away and reach the surface water [42]. engineering infrastructures. Nevertheless, it is always essential to assess the leachate hazard of C&D materials during their REFERENCES service life in pavement applications. This is reached by conducting leachate contaminant concentration [1] Tam, V.W.Y. and C.M. Tam, Crushed tests to make sure that the leachate is not putting any aggregates production from centralized negative impact on the groundwater resources and combined and individual waste sources in water streams [42]. Hong Kong. Journal of Construction and Building Materials, 2007. 21: p. 879-886. Human health concerns [2] Sustainability-Victoria, Victorian recycling In the event of using C&D material specially industries annual report 2008-2009. 2010, recycled glass in road work applications, human Sustainability Victoria: Melbourne, Australia. health risks needs to be assessed including risks to [3] Arulrajah, A., et al., Physical properties and site workers placing the material, maintenance shear strength responses of recycled workers (after development) and also incidental site construction and demolition materials in users [43]. unbound pavement base/subbase applications. The most common health concerns for FRG are Construction and Building Materials, 2014. 58: the potential for skin cuts and breathing glass dust p. 245-257. during physical handling [14]. Laboratory and field [4] Disfani, M.M., et al., Recycled crushed glass in experiences indicate that recycled glass used in this road work applications. Waste Management, research (FRG with a Dmax equal to 4.75 mm) does 2011. 31(11): p. 2341 - 2351. not harm people in the form of skin cuts and [5] Hoyos, L.R., et al., Characterization of cement punctures any more than natural aggregates like crushed rock [42]. Other researchers including fiber-treated reclaimed asphalt pavement Reference [14] mentioned that recycled glass aggregates: Preliminary investigation. Journal smaller than 9.5 mm will not cause problems unless of Materials in Civil Engineering, ASCE, 2011. it is squeezed or compressed in an ungloved hand. 23(7): p. 977 - 989. Wearing gloves during working with recycled glass [6] Poon, C.S. and D. Chan, Feasible use of eliminates the concern for skin lacerations [42]. recycled concrete aggregates and crushed clay Exposure to glass dust is another health concern brick as unbound road sub-base. Construction with recycled glass aggregate. Research outcomes and Building Materials, 2006. 20: p. 578-585. show that glass dust contains silica which in an [7] Puppala, A.J., et al., Resilient Moduli Response amorphous structure is not considered to be a health of Moderately Cement-Treated Reclaimed hazard. In any case jobsite monitoring and proper Asphalt Pavement Aggregates. Journal of personal protective equipment should be included in Materials in Civil Engineering, ASCE, 2011. any construction site safety plan [14]. 23(7): p. 990 - 998. [8] Arulrajah, A., et al., Geotechnical and geoenvironmental properties of recycled construction and demolition materials in pavement subbase applications. Journal of Materials in Civil Engineering, 2013. 25(8): p. 1077-1088.

12 12 [9] Sivakumar, V., et al., Reuse of construction [21] Taha, R., et al., Cement Stabilization of waste: performance under repeated loading. Reclaimed Asphalt Pavement Aggregate for ICE Journal of Geotechnical Engineering, Road Bases and Subbases. Journal of Materials 2004. 157(GE2): p. 91-96. in Civil Engineering, 2002. 14(3): p. 239-245. [10] Arulrajah, A., et al., Geotechnical properties of [22] Meyer, C., Recycled glass – from waste recycled concrete aggregate in pavement sub- material to valuable resource. Int. Symposium base applications. Geotechnical Testing Journal on Recycling and Reuse of Glass Cullet, , 2012. 35(5): p. 1-9. 2001(University of Dundee, Scotland). [11] Arulrajah, A., et al., Geotechnical properties of [23] Taha, B. and G. Nounu, Properties of concrete recycled crushed brick blends for pavement contains mixed colour waste recycled glass as sub-base applications. Canadian Geotechnical sand and cement replacement. Construction Journal, 2012. 49(7): p. 796-811. and Building Materials, 2008. 22(5): p. 713- [12] Arulrajah, A., et al., Geotechnical properties of 720. rwaste excavation rock in pavement sub-base [24] Wartman, J., et al., Select engineering applications. Journal of Materials in Civil characteristics of crushed glass. Journal of Engineering, 2012. 24(7): p. 924-932. Materials in Civil Engineering, 2004. 16(6): p. [13] Arulrajah, A., et al., Reclaimed asphalt 526-539. pavement/recycled concrete aggregate blends [25] Rodgers, M., et al., Cyclic loading tests on in pavement subbase applications: laboratory sandstone and limestone shale aggregates used and field evaluation. Journal of Materials in in unbound forest roads. Construction and Civil Engineering, 2014. 25(2): p. 1920-1928- Building Materials, 2009. 23: p. 2421–2427. Rahman, M.A., et al., Resilient Modulus and [26] Jitsangiam, P. and H. Nikraz, Mechanical Permanent Deformation Responses of Geogrid- behaviours of hydrated cement treated crushed Reinforced Construction and Demolition rock base as a road base material in Western Materials. Journal of Materials in Civil Australia. International Journal of Pavement Engineering, 2014. 26(3). Engineering, 2009. 10(1): p. 39 - 47. [14] Landris, T.L., Recycled glass and dredged [27] Akbulut, H. and C. Gurer, Use of aggregates materials. 2007, US Army Corps of Engineers: produced from marble quarry waste in asphalt Engineer Research and Development Center pavements. Journal of Building and Report ERDC TN-DOER-T8: Vicksburg, MS. Environment, 2007. 42: p. 1921-1930. [15] Chidiroglou, I., et al., Physical properties of [28] McKelvey, D., et al., Shear strength of recycled demolition waste material. Construction Mater, construction materials intended for use in vibro 2008. 161(3): p. 97-103. ground improvement. Ground Improvement, [16] McKinley, J.D., et al., Shear strength of Proceedings of the Institution of Civil recycled construction materials intended for Engineers, UK, 2002. 6(2): p. 59-68. use in vibro ground improvement. Ground [29] Nunes, M.C.M., et al., Assessment of Improvement, 2002. 6(2): p. 59-68. secondary materials for pavement construction: [17] Papp, W.J.J., et al., Behavior of construction Technical and environmental aspects. Waste and demolition debris in base and subbase Management, 1996. 16(1-3): p. 87-96. applications‖ Geotechnical Special Publication [30] ASTM, Standard practice for classification of (ASCE), 1998(79): p. 122-135. soils for engineering purposes (Unified Soil [18] Bennert, T., et al., Utilization of construction Classification System), in ASTM standard and demolition debris under traffic-type D2487. ASTM International. 2006: West loading in base and subbase applications. Conshohocken, Pa. Journal of Transportation Research Record, [31] ASTM, Standard Test Methods for Moisture, 2000. 1714(1350): p. 33-39. Ash, and Organic Matter of Peat and Other [19] Gabr, A.R. and D. Cameron, Properties of Organic Soils. ASTM Standard D2974. 2007, recycled concrete aggregate for unbound ASTM International: West Conshohocken, PA. pavement construction. Journal of Materials in [32] AS, Soil chemical tests—Determination of the Civil Engineering, 2012. 24: p. 754-764. pH value of a soil—Electrometric method. [20] Gregory, R.J., et al., Brick recycling and reuse. Australian Standard 1289.4.3.1. 1997, Engineering Sustainability, Proceedings of the Australian standard: Sydney, Australia. Institution of Civil Engineers UK, 2004. [33] BS, Method for Determination of Particle 157(ES3): p. 155-161. Shape; Flakiness Index. British Standard 812- 105.1. 2000, British Standard Institution: London, UK.

13 [34] VicRoads, Guide to general requirements for unbound pavement materials. Technical Bulletin 39. Vol. 39. 1998, Melbourne, VIC, Australia: VicRoads. [35] B.T., V. and R. Brimble, Austroads Repeated Load Triaxial Test Method: Determination of permanent deformation and resilient modulus characteristics of unbound granular materials under drained conditions." June by Vuong. B.T. and Brimble, R. AG-PT/T053: Reprint of APRG 00/33 (MA), 2000. [36] AS, Wastes, sediments and contaminated soils, Part 3: Preparation of leachates-bottle leaching procedure, in Australian Standards 4439.3. 1997, Standards Australia: Homebush, NSW, Australia. [37] Courard, L., et al., Use of concrete road recycled aggregates for roller compacted concrete. Journal of Construction and Building Materials, 2010. 24: p. 390-395. [38] Arulrajah, A., et al., Geotechnical performance of recycled glass-waste rock blends in footpath bases. Journal of Materials in Civil Engineering, 2013. 25(5): p. 653-661. [39] Austroads, Guide to the Structural Design of Road Pavements. 2004, Austroads: Sydney, Australia. [40] EPA, Waste Categorisation in Industrial waste resource guidelines. 2010, Environmental Protection Agency of Victoria, Australia. [41] Häkkinen, T. and S. Vares, Environmental impacts of disposable cups with special focus on the effect of material choices and end of life. Journal of Cleaner Production, 2010. 18(14): p. 1458-1463. [42] Disfani, M.M., et al., Environmental risks of using recycled crushed glass in road applications. Journal of Cleaner Production, 2012. 20(1): p. 170 - 179. [43] EPA, Solid industrial waste hazard categorization and management, industrial waste resource guidelines, Publication No. IWRG 631. 2009, Environmental Protection Agency of Victoria, Australia.

14